Model Service Settings

This page introduces Settings → Model Services the various functions on the single provider configuration page. For the complete process of configuring a provider, see Provider Configuration.

• For built-in providers, you usually only need to fill in the corresponding API key.
• Different providers call the key by different names: secret key, Key, API Key, token—they all refer to the same thing.

API key

Just enter the provider's API key. Cherry Studio supports configuring multiple keys for one provider and using them in rotation(cycling through the list from front to back). There are two ways to manage them:

• Quick Add: use English commas to separate them and enter directly, for example sk-xxx1,sk-xxx2,sk-xxx3.
• API Key Management: click the key icon button(hover to show 'API Key Management') to enter it — each key is a separate entry and can be individually edit the label,enabled / disabled,Copy,Delete.

When separating multiple keys with commas, you must use English commas.

API Address

Built-in providers generally do not require an API address (the default is fine). If you need to change it, enter it according to the provider's official documentation.

If the provider gives you https://xxx.com/v1/chat/completions a full address like this, you usually only fill in the base URL https://xxx.com and Cherry Studio will automatically append /v1/chat/completions.

If the provider's request path is not the usual /v1/chat/completions, you can enter in the API address field the full address and end it with # at the end— when it ends with # it will no longer automatically append the version path, and will only use the address you entered.

Add Model

On the provider configuration page, click the “Get model list” button to fetch the models supported by that provider; in the popped-up “Model Management” list, click the + on the right side of the model to add it to your model list.Only models added to the list will appear in model selectors throughout the app.

Below the model name, the provider-returned actual API model ID. Models with similar names, such as date versions or vendor-prefix versions, will no longer be merged; before adding or selecting a model, first check this ID to avoid choosing another version with the same name.

Connectivity Check

Click “Test” button and select a model to test whether the configuration is successful (when multiple keys are configured, you can also choose which key to test with). If the test fails, check whether the model list contains any incorrectly entered or unsupported models.

API Settings (Advanced)

Some providers can enable specific capabilities in “API Settings”, such as support for array-form message content, Developer Message,enable_thinking(controls reasoning for models like Qwen3),service_tier(OpenAI only), etc. Usually no changes are needed.

After configuring, be sure to turn on the Enable switch, otherwise the provider is still disabled and its models will not appear in the model selection list.

---

Get help and submit feedback

If you encounter any questions, bugs, or have suggestions for feature improvements during configuration or use, please refer to Feedback and Suggestions for the official channels provided.